Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

We define an inner product similar to a scalar product:

\begin{align*} \text{scalar product:}& & \vec{a}\cdot\vec{b}&= a_1b_1 + a_2 b_2 + a_3b_3 + \dots\\ \text{inner product:}& & \langle a | b \rangle &= \overline{a_1}b_1 + \overline{a_2}b_2 + \overline{a_3}b_3+\dots \end{align*}

So the inner product is designed to work in $\mathbb{C}$ similar to how scalar product works in $\mathbb{R}$. So the question arizes about a geometrical interpretation of an inner product. For scalar product it is $|\vec{a}||\vec{b}|\cos{\varphi}$ which we interpret as:

˝Projection of a 1st vector's norm to a 2nd vector multiplied by a 2nd vector's norm.˝

Question: What about for an inner product? Does the above quotation hold?

Some background on my question: I am asking this because in many quantum mechanical books they are talking of a state vector $\left|\Psi(t)\right\rangle$ which is an abstraction that defines the state of a system and is only a function of time $t$. Then we have to define a base whose base vectors and the space (we talk about position, momentum space...) we get is dependant of the base vectors we choose. So if i am talking about position space in 1-D some physicists claim that this inner product:

$$\left\langle x \right.\left|\Psi(t)\right\rangle$$

is a projection of a state vector $\left| \Psi(t)\right\rangle$ onto a normalised position base vector $\left|x\right\rangle$ and we can interpret this as $\Psi(x,t)$ which is now a function of position!

share|improve this question
5  
It's better not to use the term "scalar product" at all. The first thing is a real inner product and the second thing is a complex inner product. If you want to refer to the first thing where the $a_i$ and $b_i$ are complex, call it a symmetric bilinear form. –  Qiaochu Yuan May 26 '13 at 9:40
    
Thanks on the note! –  71GA May 26 '13 at 10:09
1  
@QiaochuYuan: unfortunately it is not that easy to kill off a French tradition. –  Willie Wong May 26 '13 at 10:58
    
Well thats how they taught us in schools. –  71GA May 26 '13 at 11:07
add comment

1 Answer

The quotation above holds in $\mathbb{C}$ as well. It might not be clear what "projection" means in the contest of complex vectors, but algebraically it is exactly as with real vectors (remember the conjugate, though).

In $\mathbb{C}$, the "inner" product of $a+ib$ and $c+id$ is: $$ ac+ bd +i (ad-bc). $$

In a way, its real part is the $\mathbb{R}^2$ scalar product, and its imaginary part is some sort of "cross product" (can you see why?). What matters, anyway, is that this way squared moduli behave exactly like in $\mathbb{R}^2$: $$ \langle(a+ib) | (a+ib)\rangle = a^2 + b^2. $$

Passing from $\mathbb{C}$ to $\mathbb{C}^n$, or any other complex vector space, pretty much the same holds: squared moduli of vectors are exactly the same as the squared moduli of $\mathbb{R}^{2n}$ vectors, while general inner product acquire a skew-symmetrical imaginary part. (This skew-symmetric part should not disturb you, because it is imaginary: think of Hermitian matrices, which generalize real symmetric matrices adding an imaginary skew-symmetric part.)

In linear algebra, you can think of the inner/scalar product as something that can "give you the coordinates". (Geometers please don't be mad at me for saying this - I agree with you but this is what they do in physics!)

For example, to get the first component of the vector $v=(x,y,z)$ we calculate its inner product with the basis element $(0,1,0)$, getting $v_2=y$.

Just as well, a function on the real line $f:x\mapsto f(x)$ is a vector in a vector space (they can be summed and multiplied by scalars). If the functions are suitably integrable (say, $L^1$ or $L^2$), we can define the inner product in the usual way. Now, a function has as its infinite components all its values $f(x)$ for all the $x$! So to know the "$x_0$-th" component, that is $f(x_0)$, we simply take the scalar product $\langle f | x_0\rangle$, where $| x_0\rangle$ is the continuous analogue of $(0,1,0)$, our "basis element". You can think of it as zero everywhere, and a sharp bump in $x_0$ (look at Dirac Delta if you don't know what it is).

Strictly speaking, the basis elements $|x\rangle$ form improper vectors, but this is a subtle, entirely different story (look at the spectral theorem, and at Schwarz distributions to know more about this).

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.